Label Encoding and One Hot Encoding in Machine Learning

📘 Label Encoding and One-Hot Encoding

Author: Bindeshwar Singh Kushwaha

---

🎯 Encoding Categorical Features

🔹 Label Encoding

• Assigns each category an integer value
• Suitable for ordinal data (e.g., size: small, medium, large)
• Tool: LabelEncoder from sklearn.preprocessing
• Example (Titanic): Encoding Sex as 0 (male), 1 (female)

🔹 One-Hot Encoding

• Converts categories into binary columns (one per category)
• Suitable for nominal data (no natural order)
• Tools: OneHotEncoder, pd.get_dummies()
• Example (Mushroom dataset): Encoding cap-shape into binary columns

---

🚢 Titanic Dataset Overview

• Source: Kaggle’s “Titanic: Machine Learning from Disaster”
• Objective: Predict survival of passengers
• Target: Survived: 0 = Did not survive, 1 = Survived
• Key Features: Pclass, Sex, Age, Fare, Embarked, etc.

---

🔢 Label Encoding in Titanic

• Assigns each category an integer
• Example (Ordinal): small → 0, medium → 1, large → 2
• Tool: LabelEncoder from sklearn
• Example on Titanic:
  • Original: ['male', 'female', 'male']
  • Encoded: [0, 1, 0]
  • Mapping: male → 0, female → 1

---

🔄 Implementation (Titanic)

import pandas as pd
from sklearn.preprocessing import LabelEncoder

df = pd.read_csv("titanic.csv")
le = LabelEncoder()
df['Sex_encoded'] = le.fit_transform(df['Sex'])

print(df['Sex'].unique())         # ['male', 'female']
print(df['Sex_encoded'].unique()) # [1, 0]
print(le.classes_)                # ['female', 'male']

---

🍄 Mushroom Dataset Overview

• Source: UCI ML Repository
• Target: class: e = edible, p = poisonous
• All Features: Categorical (e.g., cap-shape, odor, etc.)

Example: One-Hot Encoding cap-shape

• Categories: b, c, x, f, k, s
• After Encoding: binary vectors like:
  • x → [0, 0, 1, 0, 0, 0]
  • f → [0, 0, 0, 1, 0, 0]

import pandas as pd
from sklearn.preprocessing import OneHotEncoder

df = pd.read_csv("mushrooms.csv")
pd.get_dummies(df['cap-shape'], prefix='cap-shape')

---

📢 Reach PostNetwork Academy

• Website: www.postnetwork.co
• YouTube: PostNetwork Academy
• Facebook: PostNetwork Academy
• LinkedIn: PostNetwork Academy
• GitHub: PostNetwork Academy

🙏 Thank You!